WCM items not sent via syndication as expected. Event log reset explained
Occasionally the WCM Event Log Tables will become out of sync with the data in the WCM libraries. Any time content updates are not syndicated successfully to a subscriber, an event log reset should be considered.
For Portal V7 and above, there are two tasks to reset event log:
The first task is run-wcm-admin-task-reset-event-log and is sometimes referred to as the "online" event log reset. That task deletes and recreates the entries in Event Log table for that library. The server needs to be running for this task to complete. Note the task will complete in just a few seconds but the actual table rebuild is still occurring in the background. You can check the status of the background task by re-running the task from the command line without including the parameter -Dfix=true.
The second task to reset the event log is wcm-reset-event-log, aka the "offline" event log reset. That process completely drops and rebuilds the Event Log tables instead of just recreating the entries for one library. The server is down when you run the task which drops the Event Log Tables. The tables are then rebuilt when the server starts. This means it can take a long time for the server to restart, depending on the size of the wcm database and the cpu power of the portal server. For customers with a huge wcm database it can take several hours.The offline event log reset was all that was available for v 6.1.x. Starting with V7, both tasks are available. Customers should check with IBM Support before running the offline event log reset on Portal V7 and above. In some cases this task can make syndication issues worse. It is recommended to have a co-ordinated backup of the portal file system and database before running this task.
The two event log reset tasks are discussed in the Portal Infocenters here:
It is possible to run-wcm-admin-task-reset-event-log for all libraries on the base portal then all libraries on specific virtual portals. Here is sample syntax to run the task on a virtual portal.
./ConfigEngine.sh run-wcm-admin-task-reset-event-log -DallLibraries=true -Dfix=true -DVirtualPortalContext=myVirtualPortal
Note the tasks to reset the event log are run on the syndicator server only, not the subscriber (unless it is also acting as a syndicator to a third downstream server). If the syndicator is a cluster, you only need to run the tasks on the primary node.
Here are the initial steps I normally recommend customers take when items fail to syndicate:
1. Run the online event log reset for all libraries configured for the syn/sub pair
2. If no help, note the current syndicator and subscriber urls then delete and recreate the syn/sub pair
3. If no help check with IBM Support for approval to run the offline event log reset
4. If no help, recreate the syn/sub pair a second time
Also may be helpful to review the syndication troubleshooting tips found here: